New Greedy-Like Heuristics for the Multidimensional 0-1 Knapsack Problem
Richard Loulou, Eleftherios Michaelides
Abstract
Richard Loulou, Eleftherios Michaelides
Abstract
In this paper, we develop four heuristic methods to obtain approximate solutions to the multidimensional 0-1 knapsack problem. The four methods are tested on a number of problems of various sizes. The solutions are compared to the rigorous optimum as well as to a heuristic method of Toyoda. They are statistically better than the latter, with average relative errors of the order of less than 1%.
OpenAlex reports 113 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this paper, we develop four heuristic methods to obtain approximate solutions to the multidimensional 0-1 knapsack problem. The four methods are tested on a number of problems of various sizes. The solutions are compared to the rigorous optimum as well as to a heuristic method of Toyoda. They are statistically better than the latter, with average relative errors of the order of less than 1%.
Key concepts: Knapsack problem, Continuous knapsack problem, Heuristics, Greedy algorithm, Mathematical optimization, Heuristic, Change-making problem, Order (exchange)